PhD Fellowship position in Quantum-Inspired Evolutionary Algorithm for Multi-Objective Integrative Optimization
We are looking for a highly motivated individual to join as a PhD Fellow at Oslo Metropolitan University, Norway. This fully funded position focuses on Quantum-Inspired Evolutionary Algorithms. This is a fantastic opportunity to join our dynamic Artificial Intelligence research group at the Department of Computer Science and work on cutting-edge research in AI and quantum computing.
Application Deadline: August 1, 2024
The position is advertised as a 3-year position with 100% research, or a 4-year position with 75% research and 25% other tasks (teaching, supervision and/or administrative work). The goal must be to complete the PhD program/degree within the decided time frame. The decision on a 3- or 4-year position will be discussed as part of the interviews in the hiring process.
For more details about the position, including the benefits, application process and requirements, and to apply, please visit: https://www.oslomet.no/en/work/job-openings/phd-fellowship-position-in-quantum-inspired-evolutionary-algorithm-for-multi-objective-integrative-optimization
Area of research
The primary objective of this project is to formulate and implement a multi-objective quantum-inspired EAs (QEA) tailored specifically for classical computers, with a focus on addressing the prevalent challenges in the domain of multi-objective integrative optimization (MIO) problems. Real-world optimization problems, prevalent in industries, are often complex, involving different interrelated optimization problems with multiple interconnected and conflicting objectives. Most of these involved independent optimization problems are interrelated, and combining them into a global integrative optimization problem is therefore necessary. This proposal considers formulating an MIO problem by combining k optimization problems, resulting in k objective functions. As a result, instead of seeking a single solution, the approach is to provide a set of alternatives (Pareto-optimal front) that reflect the trade-off between the objectives resulting from the MIO, allowing decision-makers to choose based on their preferences. This practical approach is expected to significantly enhance the decision-making process in industries.
The MIO problems inherently fall within the NP-hard class, and traditional optimization methods often cannot handle the complexity of such real-world MIO problems, resulting in suboptimal solutions and long computational times. Evolutionary Algorithms (EAs), inspired by natural selection processes, have demonstrated effectiveness in handling NP-hard problems due to their stochastic nature, population-based exploration, and global search capabilities. However, issues like premature/slow convergence and imbalanced exploration-exploitation trade-offs limit their performance.
Recently, the emergence of quantum-inspired EAs (QEAs) has opened up new avenues for enhancing the effectiveness of EAs by striking a better balance between exploration and exploitation. Drawing inspiration from quantum mechanics, QEAs integrate concepts such as superposition, quantum parallelism, entanglement, interference, coherence, and measurement into the existing EA framework. Recent advancements have underscored the significant advantages of QEAs over classical EAs, demonstrating success in solving complex NP-hard problems that were previously deemed computationally intractable for classical computers. However, existing QEAs are typically designed for single optimization problems and exhibit optimal efficiency on specialized quantum hardware rather than classical computers. They also encounter challenges in maintaining coherence and leveraging entanglement for efficient exploration, necessitating further exploration of quantum operators and encoding schemes that can adapt to diverse problem structures and objective functions.
This project aims to bridge this gap by developing a novel multi-objective QEA that is specifically designed for classical computing environments. By utilizing quantum-inspired techniques, the objective is to provide industries with a practical and efficient solution for tackling real-world complex multi-objective optimization challenges in areas such as manufacturing and logistics. This research goal encompasses both theoretical and practical dimensions, focusing on contributing significantly to developing multi-objective QEA. Ultimately, the goal is not just to advance optimization methodologies but to facilitate broader access to cutting-edge problem-solving techniques in academic and industrial settings, thereby revolutionizing the way we approach and solve complex optimization problems.
Contact: If you would like more information about the position, feel free to contact the Principal Supervisor: Kazi Shah Nawaz Ripon, email: kazi.ripon[at]oslomet.no